PURPOSE: In general, surgical resection has been accepted as the primary treatment for resectable stage IV gastric cancer as the survival can be improved over that of non-resectional therapy, although the main prognostic factors are invasion to adjacent organs, involvement of distant lymph nodes, hepatic metastasis and peritoneal dissemination. However, there is a lack of proper criteria or surgical resection in these patients, so it is unclear which patients will benefit from a resection. METHODS: Overall, 498 patients underwent surgery, 314 had a gastrectomy, extended lymph node dissection, with or without co-resection or early postoperative intraperitoneal chemotherapy (resection group), and 184 had non-resectional surgery (non-resection group). The mean survival durations were compared with the Student's t-test. In 310 patients with a single factor, the cumulative survival rates were calculated by the Kaplan-Meier method and compared using the log-rank test. The efficacy of early postoperative intraperitoneal chemotherapy (EPIC) was also evaluated. RESULTS: In patients with one and two factors, the mean survival durations of the resection group were significantly longer than those of the non-resection group (21.4+/-23.3 months vs. 5.9+/-4.8 months; P<0.001, 13.8+/-17.5 months vs. 6.5+/-6.6 months; P=0.003, respectively). The mean survival durations with T factor were 17.4+/-18.2 and 6.3+/-4.7 months in the resection and non-resection groups, respectively (P=0.007). The mean survival durations with N factor, P factor and N&P factors were also significantly longer in resection than the non-resection group (19.2+/-21.1 months vs. 4.6+/-3.1 months; P<0.001, 26.4+/-27.4 months vs. 6.1+/-5.3 months; P<0.001, and 20.0+/-27.1 months vs. 5.7+/-5.1 months; P=0.024, respectively). The five-year survival rates of the patients with single factor were 10.9 and 0% in the resection and non-resection groups, respectively (P<0.001). The five-year survival rates of the patients with P factor were 28.9% and 9.7% in the resection and EPIC and resection only groups, respectively (P=0.0254). CONCLUSION: A survival benefit can be obtained by a surgical resection in patients with a single factor involvement, with the exception of hepatic metastasis, and with two factors involvement composed of distant lymph nodes involvement and peritoneal dissemination.